0
\$\begingroup\$

I'm no expert in high frequency. But probably it is no good idea to change the adapters sitting on type N female output on and on as a torque wrench is used to fix them.

But I have to make measurements where often a amplifier or bias tee is connected betwenn a rf-probe and a signal generator or vector network analyzer. The rf-probes have 2.92mm K connectors, the devices (bias tee, amplifier) often SMA.

Now the probe vendor says that SMA cables/connectors damage the K-connector on rf-Probe.

As I want to use the expensive low loss cables maybe also in future up to 40 GHz, currently only 1-10 GHz, I'm wondering what is the right handling here? I also want to prevent to much insertion losses by many connector adapters. The money is not really the problem of if an adapter costs 100-200€

What do you suggest here? Combining SMA and K cables with adapters or changing the test device adapters (N to SMA/K) on and on (probably 100-200 times a year? Are there adapters or products I missed that could help me?

enter image description here

\$\endgroup\$
2
  • \$\begingroup\$ Can you quantify how much insertion loss would be too much? Then measure how much you get using an adapter? \$\endgroup\$
    – Justin
    Nov 20 '19 at 19:13
  • 1
    \$\begingroup\$ @ReinstateMonica we operate with nanoelectrical physical sources of -10 dBm 10microwatt, output power of the NC in the picture. Losses of more than 3dB are not good, <1 dB would be feasible. It seems connectors loose more than low loss cables, but I'm no expert, maybe I have to research more, but some hints/directions would be good :-) \$\endgroup\$ Nov 20 '19 at 19:28
1
\$\begingroup\$

Now the probe vendor says that SMA cables/connectors damage the K-connector on rf-Probe.

Probes are very fragile and could be damaged by a very small error in operating them. They should practically be treated as consumables, and if you're using them you should budget replacing them occasionally.

So what I'd do is buy a set of probes with SMA connectors, and use them with your SMA cables.

Save the expensive K-connectorized probes (and the risk of damaging them) for the times when you need the 40 GHz bandwidth and are using K-connectorized cables.

The thing is also we have to compare measurements with same frequency response between research groups and in best case very low losses. If I have to use K-probes 1-10 GHz (they are set to 40) with max. 3 GHz test devices what do you suggest?

If you're okay with the risk your K-connector probes will be damaged, just go ahead and use them with SMA cables.

An SMA plug with nominal geometry will not damage a K jack. The risk is that the SMA standard allows looser tolerance on the center pin length, so that an in-spec SMA plug with a maximum length center pin can damage the socket in the K jack.

So, to reduce the risk, you can use new SMA cables from a reputable vendor. These are unlikely to damage your K connectors (though I can't absolutely promise they won't). If your lab has an SMA connector gauge, you can probably check which of your cables have relatively long center pins and avoid using those with the probes.

If you can't take that risk, then all you're left with is buying some SMA-K adapters and using them to connect your cables to your probes.

\$\endgroup\$
4
  • \$\begingroup\$ hmm, I didn't think of that because I didn't know when writing financial plan, but also afterwards it would be hard to argue to the funding source I need several probes with different connector to bypass my problem or save money in the long run. But I trust your answer this is probably the best to do up front.But what would you suggest to me now? The thing is also we have to compare measurements with same frequency response between research groups and in best case very low losses. If I have to use K-probes 1-10 GHz (they are set to 40) with max. 3 GHz test devices what do you suggest? Thanks \$\endgroup\$ Nov 20 '19 at 19:25
  • \$\begingroup\$ Thanks, are K-cables in general less lossy then high quality SMA cables for my frequency range of <10 GHz, or do I always have to search for vendors offering very low lossy ones? I have four K probes, so trying to use them not all with the SMA cables might also be a solution, but I'm wondering what the best low loss connection would be when testing a filter sourced by the -10dBm NC in picture and characterized by spectrum analyser tuning the filter? K-cables low loss with N/K adapter on SA? Or can I have similar loss with SMA cables and N/SMA adapter with K probes <10 GHz. Rules of thumb here? \$\endgroup\$ Nov 20 '19 at 19:42
  • \$\begingroup\$ But in the end you rather suggest to damage the probes/risk this than changing N/SMA or N/K adapters of the test devices, new and 20000 $ each? How much change of adapter with torque wrench to test devices tolerate? \$\endgroup\$ Nov 20 '19 at 19:49
  • 2
    \$\begingroup\$ @user48953094, generally as you buy cables rated for higher frequencies, they get lossier, because the inner and outer conductor diameters must be reduced to avoid moding, and this increases resistive loss. I'm not sure how big a deal this is at 40 GHz, but it is a definite problem at 70 or 80 GHz. Finding the best combination of cables and adapters is an "engineering challenge" for you. (If you're prepared to buy new cables you could also get them with SMA on one end and K on the other) \$\endgroup\$
    – The Photon
    Nov 20 '19 at 19:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.